Shave for Hope - Photo Consent & PR Risk Mitigation
💡 Core Problem
Users could upload photos of other people without consent (celebrities, ex-partners, colleagues), transform them to show a “shaved head,” and share publicly — creating serious PR and legal risk for the Childhood Cancer Foundation (CCF).
🚫 Why Camera-Only Doesn’t Work
| Constraint | Reason |
|---|---|
HTML capture attribute |
Only a hint, not enforcement. Users can dismiss camera and pick from gallery |
| No browser API blocks gallery | By design — browsers prioritize user choice |
| getUserMedia only | Excludes desktop users, requires HTTPS, spotty mobile support |
| Physical bypass | User can photograph a screen or printed photo |
The real issue is consent verification, not input method.
🔄 Risk Perspective: Why Abuse May Be Limited
| Factor | Rationale |
|---|---|
| Camera bypass is trivial | Even with camera-only, users can point the camera at a celebrity photo on another screen or print. Technical restrictions don’t prevent intent. |
| AI tools are ubiquitous | Image transformation is easily done with many public AI tools (Midjourney, Stable Diffusion, etc.). Users with malicious intent would likely use those tools instead — no friction, no consent checkbox, no accountability. |
| IG sharing is voluntary | The suggested sharing message is optional. Users may: (1) transform without sharing at all, (2) share without our message, (3) use other tools and voluntarily include our campaign. There is no direct technical binding between our app and any Instagram post. |
Implication: Bad actors have easier, anonymous alternatives. Our tool’s friction (login, consent, watermark) may actually make us a less attractive vector for abuse compared to open AI tools.
⚠️ Risk Scenarios
| Scenario | Impact | Likelihood |
|---|---|---|
| Celebrity photo transformed & shared | High — media attention, legal | Medium |
| Revenge/harassment (ex, colleague) | High — victim complaints, legal | Medium |
| Political figure mockery | High — controversy | Low-Medium |
| Stock photo abuse | Low — looks fake | Low |
🛡️ Mitigation Strategies
Tier 1: Friction-Based Deterrence (Low Effort)
| Strategy | Effect |
|---|---|
| Require login | Creates accountability trail |
| Consent checkbox | “I confirm this is my own photo” — legal cover |
| Watermark outputs | Deters sharing if watermarked |
| Disable public sharing | Keep transformations private by default |
Tier 2: Technical Detection (Medium Effort)
| Strategy | Effect |
|---|---|
| Selfie liveness check | Require blink/smile/head turn before capture |
| EXIF metadata check | Flag photos taken >24hrs ago or from different device |
| Reverse image search | Check against Google Images for public figures |
| Face comparison | If user has profile photo, compare faces match |
Tier 3: Process Controls (Operational)
| Strategy | Effect |
|---|---|
| Moderation queue | Review before allowing public share |
| Report mechanism | Let victims flag unauthorized use |
| Delayed visibility | 24hr delay before shareable |
| Terms of service | Clear liability transfer to user |
🎯 Recommendation
Minimum viable protection:
- ✅ Mandatory login — No anonymous transformations
- ✅ Consent checkbox — “I confirm I have the right to use this photo”
- ✅ Disable direct social sharing — User downloads image, shares manually
- ✅ Clear ToS — User accepts liability for misuse
- ✅ Report button — On any publicly visible transformation
If budget allows:
- 🔄 Liveness detection — Camera captures require blink/movement
- 🔄 Moderation queue — Human review before public gallery
📊 Trade-off Matrix
| Protection Level | Abuse Risk | User Friction | Viral Potential |
|---|---|---|---|
| Current (open) | High | Low | High |
| Login + consent | Medium | Medium | Medium |
| Liveness + moderation | Low | High | Low |
For a charity campaign, reputation protection > viral growth.
📝 Next Steps
- Implement mandatory login for all transformations
- Add consent checkbox with legal language to upload form
- Review and update Terms of Service
- Add report mechanism to public gallery/share pages
- Evaluate removing or gating public sharing features
- Consult with CCF legal team on liability language
🏷️ Tags Analysis
Content Analysis:
- Type:
reference(Technical decision document) - Topics:
development,web-development,business— covers technical implementation and organizational risk - Characteristics:
actionable,high-priority— requires immediate attention before launch - Priority:
high— PR crisis prevention is critical for charity reputation
Suggested Bases Filters:
- Find project notes:
project = "shave-for-hope" - Find high-priority actionable:
tags contains "actionable" AND priority = high - Find security/risk topics:
tags contains "business" AND tags contains "development"
Captured: 2025-01-13 Status: processing Next Action: Review with CCF stakeholders, implement minimum viable protection